Tracking prototype and exemplar representations in the brain across learning

Dissociable Roles of the Dorsolateral and Ventromedial Prefrontal Cortex in Human Categorization

Models of human categorization predict the prefrontal cortex (PFC) serves a central role in category learning. The dorsolateral prefrontal cortex (dlPFC) and ventromedial prefrontal cortex (vmPFC) have been implicated in categorization; however, it is unclear whether both are critical for categorization and whether they support unique functions. We administered three categorization tasks to patients with PFC lesions (mean age, 69.6 years; 5 men, 5 women) to examine how the prefrontal subregions contribute to categorization. These included a rule-based (RB) task that was solved via a unidimensional rule, an information integration (II) task that was solved by combining information from two stimulus dimensions, and a deterministic/probabilistic (DP) task with stimulus features that had varying amounts of category-predictive information. Compared with healthy comparison participants, both patient groups had impaired performance. Impairments in the dlPFC patients were largest during the RB task, whereas impairments in the vmPFC patients were largest during the DP task. A hierarchical model was fit to the participants' data to assess learning deficits in the patient groups. PFC damage was correlated with a regularization term that limited updates to attention after each trial. Our results suggest that the PFC, as a whole, is important for learning to orient attention to relevant stimulus information. The dlPFC may be especially important for rule-based learning, whereas the vmPFC may be important for focusing attention on deterministic (highly diagnostic) features and ignoring less predictive features. These results support overarching functions of the dlPFC in executive functioning and the vmPFC in value-based decision-making.

Normativity vs. uniqueness: effects of social relationship strength on neural representations of others

Understanding others involves inferring traits and intentions, a process complicated by our reliance on stereotypes and generalized information when we lack personal information. Yet, as relationships are formed, we shift toward nuanced and individualized perceptions of others. This study addresses how relationship strength influences the creation of unique or normative representations of others in key regions known to be involved in social cognition. Employing a round-robin interpersonal perception paradigm (N = 111, 20 groups of five to six people), we used functional magnetic resonance imaging to examine whether the strength of social relationships modulated the degree to which multivoxel patterns of activity that represented a specific other were similar to a normative average of all others in the study. Behaviorally, stronger social relationships were associated with more normative trait endorsements. Neural findings reveal that closer relationships lead to more unique representations in the medial prefrontal cortex and anterior insula, areas associated with mentalizing and person perception. Conversely, more generalized representations emerge in posterior regions like the posterior cingulate cortex, indicating a complex interplay between individuated and generalized processing of social information in the brain. These findings suggest that cortical regions typically associated with social cognition may compute different kinds of information when representing the distinctiveness of others.

A reduction in self-reported confidence accompanies the recall of memories distorted by prototypes

When we recall a past event, we reconstruct the event based on a combination of episodic details and semantic knowledge (e.g., prototypes). Though prototypes can impair the veracity of recall, it remains unclear whether we are metacognitively aware of the distortions they introduce. To address this, we conducted six experiments in which participants learned object-colour/object-location pairs and subsequently recalled the colour/location when cued with the object. Leveraging unsupervised machine learning algorithms, we extracted participant-specific prototypes and embedded responses in two-dimensional space to quantify prototype-based distortions in individual memory traces. Our findings reveal robust and conceptually replicable evidence to suggest that prototype-based distortion is accompanied by a reduction in self-reported confidence - an implicit measure of metacognitive awareness. Critically, we find evidence to suggest that it is prototype-based distortion of a memory trace that undermines confidence, rather than a lack of confidence biasing reconstruction towards the use of prototypes. Collectively, these findings suggest that we possess metacognitive awareness of distortions embedded in our memories.

Prototype or Exemplar Representations in the 5/5 Category Learning Task

Theories of category learning have typically focused on how the underlying category structure affects the category representations acquired by learners. However, there is limited research as to how other factors affect what representations are learned and utilized and how representations might change across the time course of learning. We used a novel "5/5" categorization task developed from the well-studied 5/4 task with the addition of one more stimulus to clarify an ambiguity in the 5/4 prototypes. We used multiple methods including computational modeling to identify whether participants categorized on the basis of exemplar or prototype representations. We found that, overall, for the stimuli we used (schematic robot-like stimuli), learning was best characterized by the use of prototypes. Most importantly, we found that relative use of prototype and exemplar strategies changed across learning, with use of exemplar representations decreasing and prototype representations increasing across blocks.

Investigating the neural effects of typicality and predictability for face and object stimuli

The brain calibrates itself based on the past stimulus diet, which makes frequently observed stimuli appear as typical (as opposed to uncommon stimuli, which appear as distinctive). Based on predictive processing theory, the brain should be more "prepared" for typical exemplars, because these contain information that has been encountered frequently, allowing it to economically represent items of that category. Thus, one could ask whether predictability and typicality of visual stimuli interact, or rather act in an additive manner. We adapted the design by Egner and colleagues (2010), who used cues to induce expectations about stimulus category (face vs. chair) occurrence during an orthogonal inversion detection task. We measured BOLD responses with fMRI in 35 participants. First, distinctive stimuli always elicited stronger responses than typical ones in all ROIs, and our whole-brain directional contrasts for the effects of typicality and distinctiveness converge with previous findings. Second and importantly, we could not replicate the interaction between category and predictability reported by Egner et al. (2010), which casts doubt on whether cueing designs are ideal to elicit reliable predictability effects. Third, likely as a consequence of the lack of predictability effects, we found no interaction between predictability and typicality in any of the four tested regions (bilateral fusiform face areas, lateral occipital complexes) when considering both categories, nor in the whole brain. We discuss the issue of replicability in neuroscience and sketch an agenda for how future studies might address the same question.

Disrupting dorsal hippocampus impairs category learning in rats

Categorization requires a balance of mechanisms that can generalize across common features and discriminate against specific details. A growing literature suggests that the hippocampus may accomplish these mechanisms by using fundamental mechanisms like pattern separation, pattern completion, and memory integration. Here, we assessed the role of the rodent dorsal hippocampus (HPC) in category learning by combining inhibitory DREADDs (Designer Receptors Exclusively Activated by Designer Drugs) and simulations using a neural network model. Using touchscreens, we trained rats to categorize distributions of visual stimuli containing black and white gratings that varied along two continuous dimensions. Inactivating the dorsal HPC impaired category learning and generalization, suggesting that the rodent HPC plays an important role during categorization. Hippocampal inactivation had no effect on a control discrimination task that used identical trial procedures as the categorization tasks, suggesting that the impairments were specific to categorization. Model simulations were conducted with variants of a neural network to assess the impact of selective deficits on category learning. The hippocampal inactivation groups were best explained by a model that injected random noise into the computation that compared the similarity between category stimuli and existing memory representations. This model is akin to a deficit in mechanisms of pattern completion, which retrieves similar memory representations using partial information.

Jackdaws form categorical prototypes based on experience with category exemplars

Categorization represents one cognitive ability fundamental to animal behavior. Grouping of elements based on perceptual or semantic features helps to reduce processing resources and facilitates appropriate behavior. Corvids master complex categorization, yet the detailed categorization learning strategies are less well understood. We trained two jackdaws on a delayed match to category paradigm using a novel, artificial stimulus type, RUBubbles. Both birds learned to differentiate between two session-unique categories following two distinct learning protocols. Categories were either introduced via central category prototypes (low variability approach) or using a subset of diverse category exemplars from which diagnostic features had to be identified (high variability approach). In both versions, the stimulus similarity relative to a central category prototype explained categorization performance best. Jackdaws consistently used a central prototype to judge category membership, regardless of whether this prototype was used to introduce distinct categories or had to be inferred from multiple exemplars. Reliance on a category prototype occurred already after experiencing only a few trials with different category exemplars. High stimulus set variability prolonged initial learning but showed no consistent beneficial effect on later generalization performance. High numbers of stimuli, their perceptual similarity, and coherent category structure resulted in a prototype-based strategy, reflecting the most adaptive, efficient, and parsimonious way to represent RUBubble categories. Thus, our birds represent a valuable comparative animal model that permits further study of category representations throughout learning in different regions of a brain producing highly cognitive behavior.

Causal Influence of Linguistic Learning on Perceptual and Conceptual Processing: A Brain-Constrained Deep Neural Network Study of Proper Names and Category Terms

Language influences cognitive and conceptual processing, but the mechanisms through which such causal effects are realized in the human brain remain unknown. Here, we use a brain-constrained deep neural network model of category formation and symbol learning and analyze the emergent model's internal mechanisms at the neural circuit level. In one set of simulations, the network was presented with similar patterns of neural activity indexing instances of objects and actions belonging to the same categories. Biologically realistic Hebbian learning led to the formation of instance-specific neurons distributed across multiple areas of the network, and, in addition, to cell assembly circuits of "shared" neurons responding to all category instances-the network correlates of conceptual categories. In two separate sets of simulations, the network learned the same patterns together with symbols for individual instances ["proper names" (PN)] or symbols related to classes of instances sharing common features ["category terms" (CT)]. Learning CT remarkably increased the number of shared neurons in the network, thereby making category representations more robust while reducing the number of neurons of instance-specific ones. In contrast, proper name learning prevented a substantial reduction of instance-specific neurons and blocked the overgrowth of category general cells. Representational similarity analysis further confirmed that the neural activity patterns of category instances became more similar to each other after category-term learning, relative to both learning with PN and without any symbols. These network-based mechanisms for concepts, PN, and CT explain why and how symbol learning changes object perception and memory, as revealed by experimental studies.

Distinct hippocampal mechanisms support concept formation and updating

Learning systems must constantly decide whether to create new representations or update existing ones. For example, a child learning that a bat is a mammal and not a bird would be best served by creating a new representation, whereas updating may be best when encountering a second similar bat. Characterizing the neural dynamics that underlie these complementary memory operations requires identifying the exact moments when each operation occurs. We address this challenge by interrogating fMRI brain activation with a computational learning model that predicts trial-by-trial when memories are created versus updated. We found distinct neural engagement in anterior hippocampus and ventral striatum for model-predicted memory create and update events during early learning. Notably, the degree of this effect in hippocampus, but not ventral striatum, significantly related to learning outcome. Hippocampus additionally showed distinct patterns of functional coactivation with ventromedial prefrontal cortex and angular gyrus during memory creation and premotor cortex during memory updating. These findings suggest that complementary memory functions, as formalized in computational learning models, underlie the rapid formation of novel conceptual knowledge, with the hippocampus and its interactions with frontoparietal circuits playing a crucial role in successful learning.

Significance statement

How do we reconcile new experiences with existing knowledge? Prominent theories suggest that novel information is either captured by creating new memories or leveraged to update existing memories, yet empirical support of how these distinct memory operations unfold during learning is limited. Here, we combine computational modeling of human learning behaviour with functional neuroimaging to identify moments of memory formation and updating and characterize their neural signatures. We find that both hippocampus and ventral striatum are distinctly engaged when memories are created versus updated; however, it is only hippocampus activation that is associated with learning outcomes. Our findings motivate a key theoretical revision that positions hippocampus is a key player in building organized memories from the earliest moments of learning.

Unraveling lexical semantics in the brain: Comparing internal, external, and hybrid language models

To explain how the human brain represents and organizes meaning, many theoretical and computational language models have been proposed over the years, varying in their underlying computational principles and in the language samples based on which they are built. However, how well they capture the neural encoding of lexical semantics remains elusive. We used representational similarity analysis (RSA) to evaluate to what extent three models of different types explained neural responses elicited by word stimuli: an External corpus-based word2vec model, an Internal free word association model, and a Hybrid ConceptNet model. Semantic networks were constructed using word relations computed in the three models and experimental stimuli were selected through a community detection procedure. The similarity patterns between language models and neural responses were compared at the community, exemplar, and word node levels to probe the potential hierarchical semantic structure. We found that semantic relations computed with the Internal model provided the closest approximation to the patterns of neural activation, whereas the External model did not capture neural responses as well. Compared with the exemplar and the node levels, community-level RSA demonstrated the broadest involvement of brain regions, engaging areas critical for semantic processing, including the angular gyrus, superior frontal gyrus and a large portion of the anterior temporal lobe. The findings highlight the multidimensional semantic organization in the brain which is better captured by Internal models sensitive to multiple modalities such as word association compared with External models trained on text corpora.

A neural network model of hippocampal contributions to category learning

In addition to its critical role in encoding individual episodes, the hippocampus is capable of extracting regularities across experiences. This ability is central to category learning, and a growing literature indicates that the hippocampus indeed makes important contributions to this form of learning. Using a neural network model that mirrors the anatomy of the hippocampus, we investigated the mechanisms by which the hippocampus may support novel category learning. We simulated three category learning paradigms and evaluated the network's ability to categorize and recognize specific exemplars in each. We found that the trisynaptic pathway within the hippocampus-connecting entorhinal cortex to dentate gyrus, CA3, and CA1-was critical for remembering exemplar-specific information, reflecting the rapid binding and pattern separation capabilities of this circuit. The monosynaptic pathway from entorhinal cortex to CA1, in contrast, specialized in detecting the regularities that define category structure across exemplars, supported by the use of distributed representations and a relatively slower learning rate. Together, the simulations provide an account of how the hippocampus and its constituent pathways support novel category learning.

A multilevel account of hippocampal function in spatial and concept learning: Bridging models of behavior and neural assemblies

A complete neuroscience requires multilevel theories that address phenomena ranging from higher-level cognitive behaviors to activities within a cell. We propose an extension to the level of mechanism approach where a computational model of cognition sits in between behavior and brain: It explains the higher-level behavior and can be decomposed into lower-level component mechanisms to provide a richer understanding of the system than any level alone. Toward this end, we decomposed a cognitive model into neuron-like units using a neural flocking approach that parallels recurrent hippocampal activity. Neural flocking coordinates units that collectively form higher-level mental constructs. The decomposed model suggested how brain-scale neural populations coordinate to form assemblies encoding concept and spatial representations and why so many neurons are needed for robust performance at the cognitive level. This multilevel explanation provides a way to understand how cognition and symbol-like representations are supported by coordinated neural populations (assemblies) formed through learning.

Rule and Exemplar-based Transfer in Category Learning

We compared the neural systems involved in transfer to novel stimuli via rule application versus exemplar processing. Participants learned a categorization task involving abstraction of a complex rule and then categorized different types of transfer stimuli without feedback. Rule stimuli used new features and therefore could only be categorized using the rule. Exemplar stimuli included only one of the features necessary to apply the rule and therefore required participants to categorize based on similarity to individual previously learned category members. Consistent and inconsistent stimuli were formed so that both the rule and feature similarity indicated the same category (consistent) or opposite categories (inconsistent). We found that all conditions eliciting rule-based transfer recruited a medial prefrontal-anterior hippocampal network associated with schematic memory. In contrast, exemplar-based transfer recruited areas of the intraparietal sulcus associated with learning and executing stimulus-category mappings along with the posterior hippocampus. These results support theories of categorization that postulate complementary learning and generalization strategies based on schematic and exemplar mechanisms.

Hierarchical Prototype Networks for Continual Graph Representation Learning

Despite significant advances in graph representation learning, little attention has been paid to the more practical continual learning scenario in which new categories of nodes (e.g., new research areas in citation networks, or new types of products in co-purchasing networks) and their associated edges are continuously emerging, causing catastrophic forgetting on previous categories. Existing methods either ignore the rich topological information or sacrifice plasticity for stability. To this end, we present Hierarchical Prototype Networks (HPNs) which extract different levels of abstract knowledge in the form of prototypes to represent the continuously expanded graphs. Specifically, we first leverage a set of Atomic Feature Extractors (AFEs) to encode both the elemental attribute information and the topological structure of the target node. Next, we develop HPNs to adaptively select relevant AFEs and represent each node with three levels of prototypes. In this way, whenever a new category of nodes is given, only the relevant AFEs and prototypes at each level will be activated and refined, while others remain uninterrupted to maintain the performance over existing nodes. Theoretically, we first demonstrate that the memory consumption of HPNs is bounded regardless of how many tasks are encountered. Then, we prove that under mild constraints, learning new tasks will not alter the prototypes matched to previous data, thereby eliminating the forgetting problem. The theoretical results are supported by experiments on five datasets, showing that HPNs not only outperform state-of-the-art baseline techniques but also consume relatively less memory. Code and datasets are available at https://github.com/QueuQ/HPNs.

Facial typicality and attractiveness reflect an ideal dimension of face structure

Face perception and recognition are important processes for social interaction and communication among humans, so understanding how faces are mentally represented and processed has major implications. At the same time, faces are just some of the many stimuli that we encounter in our everyday lives. Therefore, more general theories of how we represent objects might also apply to faces. Contemporary research on the mental representation of faces has centered on two competing theoretical frameworks that arose from more general categorization research: prototype-based face representation and exemplar-based face representation. Empirically distinguishing between these frameworks is difficult and neither one has been ruled out. In this paper, we advance this area of research in three ways. First, we introduce two additional frameworks for mental representation of categories, varying abstraction and ideal representation, which have not been applied to face perception and recognition before. Second, we fit formal computational models of all four of these theories to human perceptual judgments of the typicality and attractiveness (a strong correlate of typicality) of 100 young adult Caucasian female faces, with the models expressed within a face space derived from facial similarity judgments via multidimensional scaling. Third, we predict the perceived typicality and attractiveness of the faces using these models and compare the predictive performance of each to the empirical data. We found that of all four models, the ideal representation model provided the best account of perceived typicality and attractiveness for the present set of faces, although all models showed discrepancies from the empirical data. These findings demonstrate the relevance of mental categorization processes for representing faces.

Tackling prediction uncertainty in machine learning for healthcare

Predictive machine-learning systems often do not convey the degree of confidence in the correctness of their outputs. To prevent unsafe prediction failures from machine-learning models, the users of the systems should be aware of the general accuracy of the model and understand the degree of confidence in each individual prediction. In this Perspective, we convey the need of prediction-uncertainty metrics in healthcare applications, with a focus on radiology. We outline the sources of prediction uncertainty, discuss how to implement prediction-uncertainty metrics in applications that require zero tolerance to errors and in applications that are error-tolerant, and provide a concise framework for understanding prediction uncertainty in healthcare contexts. For machine-learning-enabled automation to substantially impact healthcare, machine-learning models with zero tolerance for false-positive or false-negative errors must be developed intentionally.

A neural network account of memory replay and knowledge consolidation

Replay can consolidate memories through offline neural reactivation related to past experiences. Category knowledge is learned across multiple experiences, and its subsequent generalization is promoted by consolidation and replay during rest and sleep. However, aspects of replay are difficult to determine from neuroimaging studies. We provided insights into category knowledge replay by simulating these processes in a neural network which approximated the roles of the human ventral visual stream and hippocampus. Generative replay, akin to imagining new category instances, facilitated generalization to new experiences. Consolidation-related replay may therefore help to prepare us for the future as much as remember the past. Generative replay was more effective in later network layers functionally similar to the lateral occipital cortex than layers corresponding to early visual cortex, drawing a distinction between neural replay and its relevance to consolidation. Category replay was most beneficial for newly acquired knowledge, suggesting replay helps us adapt to changes in our environment. Finally, we present a novel mechanism for the observation that the brain selectively consolidates weaker information, namely a reinforcement learning process in which categories were replayed according to their contribution to network performance. This reinforces the idea of consolidation-related replay as an active rather than passive process.

The effects of age on category learning and prototype- and exemplar-based generalization

The need to learn new concepts and categories persists through the lifespan, yet little is known about how aging affects the concept learning and generalization. Here, we trained young and older adults to classify typical and boundary category members, and then tested category generalization to new stimuli. During training, older adults had increased difficulty compared to young adults learning category labels for boundary items, but not typical items. At test, categorization performance that included new items at all levels of typicality was comparable across age groups, but formal categorization models indicated that older adults relied to a greater degree on generalized (prototype) category representations than young adults. These findings align with the proposal that older adults are able to form category representations based on central tendency even when they have difficulty learning and remembering individual category members. More broadly, the results contribute to our understanding of multiple categorization strategies and the limited strategy flexibility in older adults. They also highlight how reliance on preserved cognitive functions may sometimes help older adults maintain performance. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

Functional imaging analyses reveal prototype and exemplar representations in a perceptual single-category task

Similarity-based categorization can be performed by memorizing category members as exemplars or by abstracting the central tendency of the category – the prototype. In similarity-based categorization of stimuli with clearly identifiable dimensions from two categories, prototype representations were previously located in the hippocampus and the ventromedial prefrontal cortex (vmPFC) and exemplar representations in areas supporting visual memory. However, the neural implementation of exemplar and prototype representations in perceptual similarity-based categorization of single categories is unclear. To investigate these representations, we applied model-based univariate and multivariate analyses of functional imaging data from a dot-pattern paradigm-based task. Univariate prototype and exemplar representations occurred bilaterally in visual areas. Multivariate analyses additionally identified prototype representations in parietal areas and exemplar representations in the hippocampus. Bayesian analyses supported the non-presence of prototype representations in the hippocampus and the vmPFC. We additionally demonstrate that some individuals form both representation types simultaneously, probably granting flexibility in categorization strategies.

Model-based univariate and multivariate analyses of fMRI data from 62 healthy participants in a dot-pattern paradigm-based task provide further insight into the neural basis of similarity-based categorization.

Spatialization of Time in the Entorhinal-Hippocampal System

The functional role of the entorhinal-hippocampal system has been a long withstanding mystery. One key theory that has become most popular is that the entorhinal-hippocampal system represents space to facilitate navigation in one's surroundings. In this Perspective article, I introduce a novel idea that undermines the inherent uniqueness of spatial information in favor of time driving entorhinal-hippocampal activity. Specifically, by spatializing events that occur in succession (i.e., across time), the entorhinal-hippocampal system is critical for all types of cognitive representations. I back up this argument with empirical evidence that hints at a role for the entorhinal-hippocampal system in non-spatial representation, and computational models of the logarithmic compression of time in the brain.

Brain responses during auditory word recognition vary with reading ability in Chinese school-age children

While the close relationship between the brain system for speech processing and reading development is well-documented in alphabetic languages, whether and how such a link exists in children in a language without systematic grapheme-phoneme correspondence has not been directly investigated. In the present study, we measured Chinese children's brain activation during an auditory lexical decision task with functional magnetic resonance imaging. The results showed that brain areas distributed across the temporal and frontal lobes activated during spoken word recognition. In addition, the left occipitotemporal cortex (OTC) was recruited, especially under the real word condition, thus confirming the involvement of this orthographic-related area in spoken language processing in Chinese children. Importantly, activation of the left temporoparietal cortex (TPC) in response to words and pseudowords was positively correlated with children's reading ability, thus supporting the salient role phonological processing plays in Chinese reading in the developing brain. Furthermore, children with higher reading scores also increasingly recruited the left anterior OTC to make decisions on the lexical status of pseudowords, indicating that higher-skill children tend to search abstract lexical representations more deeply than lower-skill children in deciding whether spoken syllables are real. In contrast, the precuneus was more related to trial-by-trial reaction time in lower-skill children, suggesting that effort-related neural systems differ among pupils with varying reading abilities. Taken together, these findings suggest a strong link between the neural correlates of speech processing and reading ability in Chinese children, thus supporting a universal basis underlying reading development across languages.

RUBubbles as a novel tool to study categorization learning

Grouping objects into discrete categories affects how we perceive the world and represents a crucial element of cognition. Categorization is a widespread phenomenon that has been thoroughly studied. However, investigating categorization learning poses several requirements on the stimulus set in order to control which stimulus feature is used and to prevent mere stimulus-response associations or rote learning. Previous studies have used a wide variety of both naturalistic and artificial categories, the latter having several advantages such as better control and more direct manipulation of stimulus features. We developed a novel stimulus type to study categorization learning, which allows a high degree of customization at low computational costs and can thus be used to generate large stimulus sets very quickly. 'RUBubbles' are designed as visual artificial category stimuli that consist of an arbitrary number of colored spheres arranged in 3D space. They are generated using custom MATLAB code in which several stimulus parameters can be adjusted and controlled separately, such as number of spheres, position in 3D-space, sphere size, and color. Various algorithms for RUBubble generation can be combined with distinct behavioral training protocols to investigate different characteristics and strategies of categorization learning, such as prototype- vs. exemplar-based learning, different abstraction levels, or the categorization of a sensory continuum and category exceptions. All necessary MATLAB code is freely available as open-source code and can be customized or expanded depending on individual needs. RUBubble stimuli can be controlled purely programmatically or via a graphical user interface without MATLAB license or programming experience.

Learning and Representation of Hierarchical Concepts in Hippocampus and Prefrontal Cortex

A key aspect of conceptual knowledge is that it can be flexibly applied at different levels of abstraction, implying a hierarchical organization. It is yet unclear how this hierarchical structure is acquired and represented in the brain. Here we investigate the computations underlying the acquisition and representation of the hierarchical structure of conceptual knowledge in the hippocampal-prefrontal system of 32 human participants (22 females). We assessed the hierarchical nature of learning during a novel tree-like categorization task via computational model comparisons. The winning model allowed to extract and quantify estimates for accumulation and updating of hierarchical compared with single-feature-based concepts from behavior. We find that mPFC tracks accumulation of hierarchical conceptual knowledge over time, and mPFC and hippocampus both support trial-to-trial updating. As a function of those learning parameters, mPFC and hippocampus further show connectivity changes to rostro-lateral PFC, which ultimately represented the hierarchical structure of the concept in the final stages of learning. Our results suggest that mPFC and hippocampus support the integration of accumulated evidence and instantaneous updates into hierarchical concept representations in rostro-lateral PFC.SIGNIFICANCE STATEMENT A hallmark of human cognition is the flexible use of conceptual knowledge at different levels of abstraction, ranging from a coarse category level to a fine-grained subcategory level. While previous work probed the representational geometry of long-term category knowledge, it is unclear how this hierarchical structure inherent to conceptual knowledge is acquired and represented. By combining a novel hierarchical concept learning task with computational modeling of categorization behavior and concurrent fMRI, we differentiate the roles of key concept learning regions in hippocampus and PFC in learning computations and the representation of a hierarchical category structure.

How do we generalize?

Humans and animals are able to generalize or transfer information from previous experience so that they can behave appropriately in novel situations. What mechanisms-computations, representations, and neural systems-give rise to this remarkable ability? The members of this Generative Adversarial Collaboration (GAC) come from a range of academic backgrounds but are all interested in uncovering the mechanisms of generalization. We started out this GAC with the aim of arbitrating between two alternative conceptual accounts: (1) generalization stems from integration of multiple experiences into summary representations that reflect generalized knowledge, and (2) generalization is computed on-the-fly using separately stored individual memories. Across the course of this collaboration, we found that-despite using different terminology and techniques, and although some of our specific papers may provide evidence one way or the other-we in fact largely agree that both of these broad accounts (as well as several others) are likely valid. We believe that future research and theoretical synthesis across multiple lines of research is necessary to help determine the degree to which different candidate generalization mechanisms may operate simultaneously, operate on different scales, or be employed under distinct conditions. Here, as the first step, we introduce some of these candidate mechanisms and we discuss the issues currently hindering better synthesis of generalization research. Finally, we introduce some of our own research questions that have arisen over the course of this GAC, that we believe would benefit from future collaborative efforts.

Generalization and False Memory in an Acquired Equivalence Paradigm: The Influence of Physical Resemblance Across Related Episodes

The ability to make inferences about related experiences is an important function of memory that allows individuals to build generalizable knowledge. In some cases, however, making inferences may lead to false memories when individuals misremember inferred information as having been observed. One factor that is known to increase the prevalence of false memories is the physical resemblance between new and old information. The extent to which physical resemblance has parallel effects on generalization and memory for the source of inferred associations is not known. To investigate the parallels between memory generalization and false memories, we conducted three experiments using an acquired equivalence paradigm and manipulated physical resemblance between items that made up related experiences. The three experiments showed increased generalization for higher levels of resemblance. Recognition and source memory judgments revealed that high rates of generalization were not always accompanied by high rates of false memories. Thus, physical resemblance across episodes may promote generalization with or without a trade-off in terms of impeding memory specificity.

Prototype-based category learning in autism: A review

Similarity-based categorization, as an important cognitive skill, can be performed by abstracting a categories' central tendency, the so-called prototype, or by memorizing individual exemplars of a category. The flexible selection of an appropriate strategy is crucial for effective cognitive functioning. The detail-focused cognitive style in individuals with autism spectrum disorders (ASD) has been hypothesized to specifically impair prototype-based categorization but to leave exemplar-based categorization unimpaired. We first give an overview of approaches to investigate prototype-based abstraction in the prototype-distortion task, with an emphasis on model-based approaches suitable to discern the two strategies on the individual level. The second part summarizes literature speaking to prototype-based categorization in ASD using that task. Despite considerable inconsistencies, most studies appear to confirm that autistic individuals have more difficulties to perform prototype-distortion tasks than non-autistic individuals. We highlight how inconsistencies in literature can be resolved by taking the differences in task designs into account. The current review illustrates the need for sensitive computational approaches, suitable to detect hidden individual differences and potential compensatory strategies.